EP1599983B1

EP1599983B1 - Method, system and apparatus for acquiring a received impulse radio signal

Info

Publication number: EP1599983B1
Application number: EP04716736A
Authority: EP
Inventors: Andreas Molisch; Sinan Gezici; Hisashi Kobayashi; H. Vincent Poor
Original assignee: Mitsubishi Electric Corp
Current assignee: GEZICI, SINAN; Kobayashi Hisashi; POOR, H. VINCENT; Mitsubishi Electric Corp
Priority date: 2003-03-03
Filing date: 2004-03-03
Publication date: 2008-08-20
Anticipated expiration: 2024-03-03
Also published as: WO2004079921A3; US20040176120A1; JP2006521760A; DE602004015941D1; EP1599983A2; JP4335912B2; US7292619B2; WO2004079921A2

Description

Technical Field

The invention relates generally to impulse radio signals, and more particularly to acquiring ultra-wide-bandwidth signals.

Background of the Invention

Before a received ultra-wide-bandwidth (UWB) signal can be demodulated, a template signal must be aligned with the received signal. The purpose of the alignment is to determine a relative delay of the received signal with respect to the template signal. This process is called signal acquisition.
Conventionally, the alignment is performed by a serial search of possible delay times of cells in an uncertainty region, see Simon et al., "Spread Spectrum Communications Handbook," McGraw-Hill, New York, 1994. Each different search location, i.e., time interval, with respect to time, is called a cell. If the signal exists at a delay location, then that cell is called a signal cell. In that method, the received signal is correlated with the template signal, and an output of the correlation is compared to a threshold. If the output is lower than the threshold, then the template signal is shifted by some amount time. The shifted amount time corresponds usually to a resolvable path interval. This information is then used to repeat the correlation until the output exceeds the threshold.
If the output of the correlation comes from a case where the signal path and the template signal are aligned, it is called a signal cell output, otherwise, it is called a non-signal cell output. A false alarm occurs when a non-signal cell output exceeds the threshold. In that case, time tp elapses until the search recovers. This time is called the penalty time for false alarm.
Due to the short time resolution of UWB signals, serially searching all delay locations can take a long time. Therefore, the alignment method must be fast so that the time allocated for the acquisition of the UWB signal is reduced.
The mean acquisition time of a serial search is directly proportional to the number of cells in the uncertainty region, N, for large N values, see Polydoros et al., "A unified approach to serial search spread-spectrum code acquisition-Part I: General Theory," IEEE Trans. Comm., Vol. COM-32, pp. 542-549, May 1984. In a general form, Figure 3 shows the basic operation of the prior art serial cell search of an uncertainty region 300 with a template 301.
WO-A-96/41432 discloses a method for acquiring a received impulse radio signal.
Therefore, there is a need for a method and system that can acquire UWB signals in a shorter time than the known prior art serial search techniques.

Disclosure of the Invention

In impulse radio communication systems, such as ultra-wide-bandwidth (UWB) communications, the received radio signal must be acquired before the signal can be demodulated. Due to the very short time resolution of the narrow pulses of the UWB signal, a considerable number of possible signal locations, i.e., delays, must be searched in order to acquire the received signal. For this reason, fast signal acquisition is very important for impulse radio communications.
Using a first template signal, a sequential block search method (SBS) according to the invention first determines a small region, with respect to time,
where the received signal cells are likely to exist. Then, this region is searched in more detail to find the exact delay of the signal using a second template signal.
The correlation of the received signal with the first template signal effectively adds some cell outputs. The result is then used as a criterion to determine if that region, generally, contains signal cells.
If the correlated output for the region exceeds a specified threshold value, then the block is searched serially in detail using the second template signal.
The correlation intervals for the sequential block search and the serial cell search steps do not have to be the same. In fact, the correlation interval of the block search is usually chosen to be substantially longer in order to be able find the correct block with a high probability.
In harsh non-line-of-sight (NLOS) conditions, there are many multipaths, which make signal acquisition more challenging. If the first path, or first few paths, need to be acquired in such conditions, then an average block search method according to the invention can be used. In this method, a number of serial search outputs are averaged, and the amount of increase between successive average values is compared to a threshold to detect an edge of the start of the received signal.
If there is a significant increase in the average value, then the outputs of the cells causing that increase are further searched to determine the first path of the received signal.

Brief Description of the Drawings

Figure 1 is a block diagram of a sequential block search method and system according to the invention; and
Figure 2 is a block diagram of an average block search method and system according to the invention.

Best Mode for Carrying Out the Invention

Signal Model

In a binary phase-shift keyed random time hopping impulse radio (TH-IR) system, the transmitted signal can be represented by the following model: $S_{tr} (t) = \sum_{j = - \infty}^{\infty} d_{j}^{k} b_{⌊ j / N_{f} ⌋}^{k} w_{tr} (t - j T_{f} - c_{j}^{k} T_{c}),$

where W_tr is the transmitted unit-energy pulse, T_f is the average pulse repetition time, N_f is the number of pulses representing one information symbol, and b is the information symbol transmitted, i.e., zero or one.
In order to allow the channel to be exploited by many users and avoid catastrophic collisions, a pseudo-random sequence {c_j} is assigned to each user. This sequence is called the time hopping (TH) sequence. The TH sequence provides an additional time shift of C_j T_c seconds to the j^th pulse of the signal, where T_c is sometimes called the chip interval. To prevent pulses from overlapping, the chip interval is selected to satisfy T_c ≤ T_f /N_c.
We consider coded IR systems where the d_j's are binary random variables, and where d_i and d_j are independent for i ≠ j, taking each values ± 1 with a probability of ½, see Fishler et al., "On the tradeoff between two types of processing gain," 40th Annual Allerton Conference on Communication, Control, and Computing, 2002. This systems can be regarded as an random - code division multiple access radio signal (RCDMA) system with T_f = T_c. In this case, N_f represents the processing gain.
We define a sequence {S_j} as follows $sj = {\begin{matrix} d_{⌊ j / N_{c} ⌋} & j - N_{f} ⌊ j N_{c} ⌋ = c_{⌊ j {/ N}_{c} ⌋} \\ 0 & otherwise \end{matrix} .$
Then, assuming T_f / T_c = N_c, without loss of generality, Equation (1) can be expressed $S_{tr} (t) = \sum_{j = - \infty}^{\infty} s_{j} b_{⌊ j / N_{f} N_{c} ⌋}^{k} w_{tr} (t - j T_{c}) .$
We assume that no data modulation is done during the acquisition stage, that is b ^k _j/NfNc = 1∀j. In this case, the received signal over a flat fading channel in a single user system can be expressed as $r (t) = \sum_{j = - \infty}^{\infty} s_{j} w_{rec} (t - j T_{c} - τ) + σ_{n} n (t),$

where W_rec(t) is the received UWB pulse, and n(t) is white Gaussian noise with unit power spectral density. This model approximately represents the line-of-sight (LOS) case, with a strong first component.
The number of cells in an uncertainty region is taken to be N = N_f N_c. One of these cells is the signal cell, while the others are non-signal cells.

Template Signal

Assuming no data modulation for the purposes of acquisition, then the template signal that is used in a serial search for the signal model in Equation (3) can be expressed as follows: $S_{m_{2}}^{(c)} (t) = \sum_{m = {jN}_{c}}^{(j + m_{2}) N_{c} - 1} s_{n} w_{rec} (t - n T_{c}),$

where m₂ is the number of pulses, over which the correlation is taken.

Sequential Block Search

For a sequential block search (SBS) according to the invention, there are two different template signals. The first template signal is used for searching a block of cells, while the second template signal is similar to the one used in the serial search.
The first template signal for the signal model described in Equation (3) can be expressed as follows: $S_{m_{1}}^{(b)} (t) = \sum_{i = 0}^{K - 1} \sum_{m = {jN}_{c}}^{(j + m_{1}) N_{c} - 1} s_{n} w_{rec} (1 - {nT}_{c} - {iT}_{c} - (b - 1) K T_{c}),$

where B is the total number of blocks in the uncertainty region, each block including K cells, and where m₁ is the number of pulses, over which the correlation is taken. For simplicity, it is assumed that the total number of uncertainty cells can be expressed as N = KB. The value T_c is taken as the minimum resolvable path interval.
The output of the correlation of the received signal and the first template signal in Equation (6) is used as a quick test to check if the whole block contains a signal cell, or not. The correlation output of the received signal and the second template signal is then used in a detailed search of a block.
The index of the block that is currently being searched is b, with b = 1 initially. Then, the SBS method can be described as follows:

1) Check the b^th block using the first template signal S ^(b) _m1 (t).
2) If the output of the b^th block is not higher than a block threshold, τ_b, then, go to step 6.
3) If the output of the b^th block is higher than the block threshold, τ_b, then search the block in more detail, i.e., cell-by-cell serial search with a signal
threshold τ_s, using the second template signal S_m2 ^(c) (t).
4) If no signal cell is detected in the block, go to step 6.
5) If the signal cell is detected in the block, DONE.
6) Set b = (b mod B) + 1 and go to step 1.

When a false alarm (FA) occurs in the serial search part, the search resumes with the next cell after C time units, which is the penalty time in terms of frame time.
In step 5, "the signal cell is detected" means that the signal cell output exceeds the signal threshold, τ_s. Similarly, in step 4, "no signal cell is detected" implies that the signal cell is not in the block, or the output of the cell is lower than the signal threshold τ_s, even if the cell is in the block.
Figure 1 shows the SBS method. The received signal 101 is correlated 110 with the first template signal of Equation (6), and the output 111 is compared 120 to the block threshold ^τb.
If the block threshold is not exceeded 121, the decision unit has a synchronization unit 130 adjusted 131 the delay of the first template signal, and another correlation 110 with the received signal is performed.
When the block output 111 is higher than the block threshold ^τb, the second template signal in Equation (5) is employed and the cells in the block are serially searched. In other words, decision unit 120 compares the outputs with the thresholds and decides if the signal is detected 122, or not 121, while the synchronization unit 130 adjusts 131 the delays of the template signals and sends the corresponding one to the correlation unit.

Average Block Search

An average block search method is appropriate in harsh NLOS conditions. The basic idea behind this method is to use an average value of a number of serial correlation outputs in order to see a considerable increase in the output values. This increase indicates the start of the signal cells. The received signal in this case is expressed as: $r (t) = \sum_{j = \infty}^{\infty} \sum_{l = 1}^{L} α_{i} s_{j} w_{rec} (t - j T_{c} - τ_{l}) + σ_{n} n (t),$

where ^α1 is the amplitude coefficient and ^τ1 is the delay of the 1^th multipath component.
Consider the outputs of the correlations of the received signal with the following template signal: $S_{m}^{(c)} (t) = \sum_{m = {jN}_{c}}^{(j + m) N_{c} - 1} s_{c} w_{rec} (t - n T_{c}) .$
If the absolute values of the results of these correlations are z₁,...,z_N, then we can define $w_{i} = \frac{1}{K} \sum_{j = iK + 1}^{(i + 1) K} z_{j},$

assuming N = KB.
Let i be the index of the averaged block currently being searched, with i = 0 initially. Then, the ABS method can be described follows:

1) Check difference between successive averages w_{i mod B} - w_(i-1) mod B.
2) If the difference is not higher than a first threshold ^τa go to step 6.
3) If the difference is higher than ^τa, check Z_{(i mod B)K+1}, ...,Z_{(i mod B)+1)K}
serially, comparing to a second thresholds, ^τc.
4) If no signal cells detected, go to step 6.
5) If signal cell(s) are detected, DONE.
6) Set i = (i + 1) mod B, and go to step 1.

Figure 2 shows ABS method and system 200. In this embodiment multiple correlators 210 averaging units 215 are used in parallel. A received signal r(t) 201 is first correlated 210 with a first template signals with different delays. Then, the absolute values of these correlations are averaged 220 and compared to the previous averaged value by the decision unit 230. If there is a significant increase in the average value and if any one of the serial search outputs in the corresponding block exceeds the threshold, the signal is detected 231.
If no detection 232 occurs, then, the delays of the template signal are adjusted by the synchronization unit 240, and the same steps are followed again.
Note that even though the block diagram is shown for the case with K correlators and averaging units, the method and system can also be worked with only one correlator. In such a situation, the decision unit can perform the averaging and comparison tasks by storing a predetermined number of outputs of the single correlator.

Effect of the Invention

The sequential block search method according to the invention provides a quick method to find the location(s) of a signal cell of a UWB signal. First, the method quickly determines a smaller region where signal cells are likely to exist. Then, it searches that region in detail to find the exact location of the signal. In this way, the time to acquire the UWB signal can be reduced considerably. In fact, the mean acquisition time of the SBS method becomes proportional to the square root of N for large signal-to-noise ratios. In contrast, the mean acquisition time of a prior art serial search is directly proportional to the number of cells in an uncertainty region. For practical values, the acquisition time using the SBS method is about the half of the serial search mean acquisition time.
In harsh multipath conditions, an average block search reduces the acquisition time because the averaged values of serial search outputs are more reliable in detecting the start of the signal in some NLOS situations. In this way, instantaneous increases in the single outputs are smoothed so that the frequency of false alarms is reduced. It should be noted that the invention can also be used in direct sequence - code division multiple access (DS-CDMA) systems.

Claims

A method for acquiring a received impulse radio signal, comprising:
searching, with respect to time, a region of the impulse radio signal using a first template signal to locate a block including a signal cell; and

searching, with respect to time, the block using a second template signal to locate the signal cell to acquire the received impulse radio signal.
The method of claim 1 wherein the impulse radio signal is a random - code division multiple access radio signal.
The method of claim 1 wherein the impulse radio signal is an ultra-wide-bandwidth radio signal.
The method of claim 1 wherein the impulse radio signal is a direct sequence - code division multiple access radio signal.
The method of claim 1 wherein the impulse radio signal is a direct binary phase-shift keyed random time hopping radio signal.
The method of claim 1 further comprising:
wherein a first correlation interval associated with the first template signal is substantially longer than a second correlation interval associated with the second template signal.
The method of claim 5 wherein the searching further comprises:
time shifting the first and second template signals while searching;

correlating the shifted first and second template signals over the first and second correlation intervals to determine first and second outputs, respectively; and

comparing the first and second outputs to first and second thresholds to locate the block and signal cell, respectively.
The method of claim 1 wherein a transmitted impulse radio signal corresponding to the received impulse radio signal is represented by $S_{tr} (t) = \sum_{j = - \infty}^{\infty} d_{j}^{k} b_{⌊ j / N_{f} ⌋}^{k} w_{tr} (t - j T_{f} - c_{j}^{k} T_{c}),$

where w_tr is a transmitted unit-energy pulse, T_f is an average pulse repetition time, N_f is the number of pulses representing one information symbol, and b is an information symbol transmitted, either zero or one, and a pseudo-random sequence {c_j} is assigned to each user of the impulse radio signal to provide an additional time shift of c_j T_c seconds to the j^th pulse of the impulse radio signal, where Tc is a chip interval, and T_c ≤ T_f /N_c, and where d_j are binary random variables, and where d_i and d_j are independent and taking each values of ± 1 with a probability of ½, and a sequence {s_j} as defined follows by $s_{j} = {\begin{matrix} d_{⌊ j / N_{c} ⌋} & j - N ⌊ j N_{c} ⌋ = c_{⌊ j / N_{c} ⌋} \\ 0 & otherwise \end{matrix},$

to express the transmitted impulse radio signal as $S_{tr} (t) = \sum_{j = - \infty}^{\infty} s_{j} b_{⌊ j / N_{f} N_{c} ⌋}^{k} w_{tr} (t - j T_{c})$

when T_f / T_c = N_c, so that the received impulse radio signal is expressed as $r (t) = \sum_{j = - \infty}^{\infty} s_{j} w_{rec} (t - j T_{c} - τ) + σ_{n} n (t),$

where w_rec(t) is a received pulse, and σ_nn(t) is white Gaussian noise with unit power spectral density, and b^k _jlnjnt, and the number cells in the region is N = N_fN_c,, and T_c is a minimum resolvable path interval.
The method of claim 8 wherein the first template signal is expressed $S_{m_{1}}^{(b)} (t) = \sum_{i = 0}^{K - 1} \sum_{m = {jN}_{c}}^{(j + m_{1}) N_{c} - 1} s_{n} w_{rec} (1 - {nT}_{c} - {iT}_{c} - (b - 1) K T_{c}),$

where m₁ is the number of pulses over a first correlation interval associated with the region, B is a total number of blocks in the region, each block including K cells, and the plurality of cells is N = KB.
The method of claim 8 wherein the second template signal is expressed as $S_{m_{2}}^{(c)} (t) = \sum_{m = {jN}_{c}}^{(j + m_{2}) N_{c} - 1} s_{n} w_{rec} (t - n T_{c}),$

where m₂ is the number of pulses over a second correlation interval associated with the block.
The method of claim 1 wherein the searching of the region uses a block threshold, τ_b, and the searching of the block uses a signal threshold τ_s.
A system for acquiring a received impulse radio signal, comprising:
first means for searching, with respect to time, a region of the impulse radio signal using a first template signal to locate a block including a signal cell; and

second means for searching, with respect to time, the block using a second template signal to locate the signal cell to acquire the received impulse radio signal.
An apparatus for acquiring a received impulse radio signal, comprising:
a correlation unit configured to search, with respect to time, a region of the impulse radio signal using a first template signal to locate a block including a signal cell, and to search, with respect to time, the block using a second template signal to locate the signal cell to acquire the received impulse radio signal;

a synchronization unit to time-shift the first and second template signals while searching; and

a decision unit to compare first and second outputs of the correlation unit to first and second thresholds while searching the region and block respectively.